SCSI related hang on boot

DISCUSSION Ron Day 2 participants 2 messages Jun 20, 2004

Share & tools

Ron Day

author

2004-06-20 16:11 UTC #0 lore raw

Please CC me on this, since I'm not a list member.

I've been following this thread and hoping for an answer. The problem I'm having, as well as my hardware, is very similar to this.

For the record, my equipment:

ASUS A7M-266D, 2 x Athlon MP1800+, 1024MB ECC registered

Tekram DC-390U3W, Symbios Logic (LSI) 53c1010 chipset.
Running for 2+ years with sym53c8xx_2 driver. It replaced a Fireport40 that ran for 2-3 years with the same driver and the earlier sym53c8xx (v1) with older/slower/smaller HDD's.

Bus0, terminated in-line after second HDD
2 x U160 HDD's, id 0 and 1 (2+ years running)

Bus1, terminated in-line at scanner
An internal Yamaha SCSI CD-RW, id 2 (3-4 years running) A UMAX S-12 scanner, id 5 (8+ years running)

Everything goes as expected through sym0, where the two HDD's reside, but problems arise at sym1. Compared to successful dmesg output, it seems that the scanner (and hence the termination for that bus) is not seen.

My conclusion is that sym2.1.18i works and sym2.1.18j does not. I say this because I have made a copy of the drivers/scsi/sym53c8xx_2 directory from 2.6.5 (the latest that I'm aware of with the .i drivers). By replacing the files in current kernels with these, I have compiled and run several newer versions sucessfully, the latest being 2.6.7, which I'm running now.

Rather than post everything here, I'll link to some files that should be relevant.

The error (written down and typed in, no serial console):
http://ronmon.shacknet.nu/configs/scsi_boot_error

Output from a good dmesg (sym2.1.18.i driver):
http://ronmon.shacknet.nu/configs/dmesg_good-2.6.7-rc3

Brief system layout courtesy of phpsysinfo:
http://ronmon.shacknet.nu/phpsysinfo/

Output from 'lspci -vvx':
http://ronmon.shacknet.nu/configs/lspci-vvx

Kai OM

2004-06-20 21:50 UTC #1 lore raw

I've got an LSIU160 controller, which means that it has two actual controllers, an exterior and an interior one.
The exterior one is disabled in the controller's BIOS, because I have no use for it.
The interior on is terminated immediately after the drive I have connected to it.
I'm wondering if the driver is somehow ignoring(at least in my case) fact that the disabled controller is, in fact, disabled, and is getting thrown off because I never bothered to buy a terminator for it, when it's not supposed to ever be active.
I'm not a SCSI expert, though, so I'm prolly wrong.

----- Original message -----
From: "Ron Day" <[email protected]>
To: [email protected]
Date: Sun, 20 Jun 2004 12:11:01 -0400
Subject: Re: SCSI related hang on boot